PROJECT 1: ENCODING AND RETRIEVAL PROCESSES IN DECLARATIVE MEMORY In the prior funding period, patients with and at risk for schizophrenia were found to show a differential deficit in recollection that scaled with phase of illness, while familiarity was intact. This deficit was correlated with reduced hippocampal and parahippocampal activity at retrieval on items for which associative memory was evident. There are two (not mutually exclusive) possible sources ofthese deficits: 1) disrupted structural connectivity, resulting from aberrant developmental changes in gray and/or white matter during peri-adolescence; and 2) failure of patients to disengage default mode network (DMN) activity during memory encoding and/or retrieval. The first hypothesis is suggested by our preliminary work indicating a steeper rate of gray matter reduction in prefrontal regions and a lack of normal age-related increase in medial temporal white matter tracks in FE and prodromal patients who convert to psychosis. The second hypothesis is based on recent evidence of reduced suppression of DMN activity and less task-related activity and functional connectivity during a working memory task in schizophrenia patients and their siblings. Increased DMN activation appears to accompany attention to internal thoughts and self-referential thinking; reduced suppression ofthe DMN may thus interfere with patients' ability to activate the episodic and associative memory circuitry at encoding and/or retrieval. In the next phase of the study we will pursue these two hypotheses in cross-sectional and short-term longitudinal studies of prodromal, FE and chronic patients and controls, using MRI and DTI to index structural connectivity and fMRI during an associative memory task and resting scans to index functional connectivity between memory-related and DMN regions. The findings will help to clarify the nature and neural underpinnings of declarative memory deficits in schizophrenia, whether behavioral and physiologic deficits predict the onset of schizophrenia over and above behavioral indicators of risk and are differentially predictive of poor functional outcome. Findings maybe used in prodromal research to identify those at greatest risk for schizophrenia.